terraform-provider-helm: Error: unable to build kubernetes objects from release manifest: unknown

Terraform Version and Provider Version

Terraform version: 0.12.24

Provider Version

provider.helm v1.2.1

Affected Resource(s)

provider
helm_release

Terraform Configuration Files

provider "helm" {
  kubernetes {
    host     = var.openshift_server
    username = var.openshift_admin
    password = var.openshift_password
  }
}
resource "helm_release" "main" {
  name = "sumologic"
  repository = "https://kubernetes-charts.storage.googleapis.com/"
  chart     = "sumologic-fluentd"
  namespace = "logging"

  values = [
    "${file("${path.module}/values.yml")}"
  ]
}

Debug Output

12:47:33 [DEBUG] Installing Chart
2020-05-29T12:47:33.101-1000 [DEBUG] plugin.terraform-provider-helm_v1.2.1_x4: 2020/05/29 12:47:33 [INFO] Successfully initialized config
2020/05/29 12:47:34 [DEBUG] module.sumo_test.helm_release.main: apply errored, but we're indicating that via the Error pointer rather than returning it: unable to build kubernetes objects from release manifest: unknown
2020/05/29 12:47:34 [ERROR] module.sumo_test: eval: *terraform.EvalApplyPost, err: unable to build kubernetes objects from release manifest: unknown
2020/05/29 12:47:34 [ERROR] module.sumo_test: eval: *terraform.EvalSequence, err: unable to build kubernetes objects from release manifest: unknown

Error: unable to build kubernetes objects from release manifest: unknown

  on ../main.tf line 8, in resource "helm_release" "main":
   8: resource "helm_release" "main" {


2020-05-29T12:47:34.492-1000 [DEBUG] plugin: plugin process exited: path=/helm_charts/sumo-fluentd/example/.terraform/plugins/darwin_amd64/terraform-provider-helm_v1.2.1_x4 pid=71709

Expected Behavior

fluentd helm chart is deployed

Actual Behavior

Error: unable to build kubernetes objects from release manifest: unknown

Steps to Reproduce

terraform apply

About this issue

Original URL
State: closed
Created 4 years ago
Reactions: 6
Comments: 18

Most upvoted comments

I’ve got this weird error when I had the wrong authorization in the provider. This patch work for me. Screenshot 2021-08-18 at 17 02 53

+16

ondrejsika on Aug 18, 2021

In my case this error reproduced because I was using v1 of the Kubernetes provider and v2 of the Helm provider. Changing both to the same version (in my case v1) resolved the issue.

MikiLoz92 on Mar 3, 2021

I’m facing same issue. I’m using latest helm provider and deploying the latest bitnami nginx chart. I was able to deploy it to the same cluster with helm install from my workstation. but when running from terraform ( also on my local machine) , I’m getting This error Error: unable to build kubernetes objects from release manifest: unknown

msldn on Jul 11, 2020

@MikiLoz92 i can confirm that when i upgrade both provider to version 2.x.x it seems to work well. Before that i had kubernetes provider 1.x.x and helm provider 2.x.x Thanks!

rtrive on Mar 24, 2021

Same here. Something is causing the GKE masters to cascade crash, rendering the whole cluster unmanageable for a while until they recover. Happens with both v1.3.x and v2.x of the provider.

athak on Feb 17, 2021

Same behaviour, but at least in my case, it was my fault.

I was creating the k8s cluster with this config (GKE with google_container_cluster resource):

resource "google_container_cluster" "xxx" {

    master_auth {
        client_certificate_config 
        {  
            issue_client_certificate = false  
        }  
    }

So the k8s state didn’t contain client certificates 😃

sergiorc on Aug 26, 2020